When Türkiye implemented regional development subsidies in 2012, traditional evaluation showed modest success among targeted firms in less developed regions. However, comprehensive analysis using firm-to-firm transaction data revealed that suppliers and customers outside the targeted areas also experienced significant performance gains. The policy achieved far higher economic impact than measured.
This pattern emerges consistently across countries and policy types. In Estonia, increases in the price of carbon (from the EU Emissions Trading System) generated energy efficiency improvements not just among directly targeted firms, but also among their business partners who received no carbon price signal. Belgian innovation policies created productivity spillovers through supply chains rather than geographic proximity. Japanese technology investment incentives boosted the sales of suppliers of targeted firms while leaving customers unaffected.
The common thread: industrial policies operate within complex networks, where effects cascade far beyond direct recipients. Yet, traditional evaluation systems only capture the tip of the iceberg.
Why effects are systematically underestimated and why it matters
For example, when governments provide Company A with incentives to adopt digital technologies, Companies B, C, and D within their supply network frequently implement similar technologies—not through direct government intervention, but via business relationships that facilitate knowledge transfer, create competitive pressures, and enable technology diffusion.
OECD governments are currently deploying unprecedented resources toward economic transformation. If evaluation systems capture only part of the actual impact, policy decisions are being made with fundamentally incomplete information about what works, what doesn't, and where to allocate limited public resources for maximum effect.
How can we measure industrial policies more effectively?
If these network effects are so significant, why have they remained invisible to policymakers? The answer is simple: there was no readily available data to systematically link firms. And without these links, there is no way to identify networks and assess network effects.
However, every month, millions of firms submit detailed sales and purchase reports to tax authorities, value-added tax (VAT) data, creating comprehensive, real-time maps of commercial relationships that reveal exactly who sells to whom, for how much, and how these relationships evolve.
This administrative data provides unprecedented insights into economic networks, revealing granular firm-level connections, precise pathways through which policy effects propagate, and highly connected firms whose behavioral changes cascade throughout the entire system.
How can governments better leverage VAT data
Currently, most VAT data remain locked within tax administrations due to confidentiality regulations, institutional silos between tax and policy departments, and lack of recognition that transaction data represents essential policy infrastructure. However, countries such as Estonia have demonstrated that secure analytical environments can enable policy research while maintaining privacy protections.
There are two key approaches to leveraging this data and the insights it can offer:
- Establish data access infrastructure that enables cross-agency partnerships between tax administrations, statistical offices, and policy departments, supported by legal frameworks that authorise policy research while protecting firm information.
- Build analytical capabilities in network analysis techniques for mapping business relationships, spillover identification methods, and evaluation frameworks that capture both direct and indirect impacts.
The utility of VAT data can even extend beyond better understanding and measuring the impact of industrial policies. For example, just as firms may adopt the digital technologies of a supplier, unforeseen disruptions to an upstream supplier may spread to downstream buyers. In this context of negative spillovers, VAT data could allow policymakers to map the entire supply chain and identify policies to enhance supply chain resilience.
How policymakers can engage on the topic of VAT data
To assist policymakers in leveraging VAT data, a new OECD report on transaction data for evidence-based industrial policy provides detailed technical guidance and country case studies for implementing network-aware policy making. The OECD, in partnership with DG GROW, has also created the Leveraging Inter-Firm Transaction (LIFT) network to facilitate VAT data access and leverage VAT data for policy analysis.
Industrial policies operate within complex networks, where effects cascade far beyond direct recipients. The analytical tools and data to understand these effects exist within government systems. What remains is the institutional commitment to unlock this data and transform policy design to account for economic reality.
